linux/crypto/ciphers/des/VERSION

   1 Version 4.04
   2         Fixed a few tests in destest.  Also added x86 assember for
   3         des_ncbc_encrypt() which is the standard cbc mode function.
   4         This makes a very very large performace difference.
   5         Ariel Glenn ariel@columbia.edu reports that the terminal
   6         'turn echo off' can return (errno == EINVAL) under solaris
   7         when redirection is used.  So I now catch that as well as ENOTTY.
   8
   9
  10 Version 4.03
  11         Left a static out of enc_write.c, which caused to buffer to be
  12         continiously malloc()ed.  Does anyone use these functions?  I keep
  13         on feeling like removing them since I only had these in there
  14         for a version of kerberised login.  Anyway, this was pointed out
  15         by Theo de Raadt <deraadt@cvs.openbsd.org>
  16         The 'n' bit ofb code was wrong, it was not shifting the shift
  17         register. It worked correctly for n == 64.  Thanks to
  18         Gigi Ankeny <Gigi.Ankeny@Eng.Sun.COM> for pointing this one out.
  19
  20 Version 4.02
  21         I was doing 'if (memcmp(weak_keys[i],key,sizeof(key)) == 0)'
  22         when checking for weak keys which is wrong :-(, pointed out by
  23         Markus F.X.J. Oberhumer <markus.oberhumer@jk.uni-linz.ac.at>.
  24
  25 Version 4.01
  26         Even faster inner loop in the DES assembler for x86 and a modification
  27         for IP/FP which is faster on x86.  Both of these changes are
  28         from Svend Olaf Mikkelsen <svolaf@inet.uni-c.dk>.  His
  29         changes make the assembler run %40 faster on a pentium.  This is just
  30         a case of getting the instruction sequence 'just right'.
  31         All credit to 'Svend' :-)
  32         Quite a few special x86 'make' targets.
  33         A libdes-l (lite) distribution.
  34
  35 Version 4.00
  36         After a bit of a pause, I'll up the major version number since this
  37         is mostly a performace release.  I've added x86 assembler and
  38         added more options for performance.  A %28 speedup for gcc
  39         on a pentium and the assembler is a %50 speedup.
  40         MIPS CPU's, sparc and Alpha are the main CPU's with speedups.
  41         Run des_opts to work out which options should be used.
  42         DES_RISC1/DES_RISC2 use alternative inner loops which use
  43         more registers but should give speedups on any CPU that does
  44         dual issue (pentium).  DES_UNROLL unrolls the inner loop,
  45         which costs in code size.
  46
  47 Version 3.26
  48         I've finally removed one of the shifts in D_ENCRYPT.  This
  49         meant I've changed the des_SPtrans table (spr.h), the set_key()
  50         function and some things in des_enc.c.  This has definitly
  51         made things faster :-).  I've known about this one for some
  52         time but I've been too lazy to follow it up :-).
  53         Noticed that in the D_ENCRYPT() macro, we can just do L^=(..)^(..)^..
  54         instead of L^=((..)|(..)|(..)..  This should save a register at
  55         least.
  56         Assember for x86.  The file to replace is des_enc.c, which is replaced
  57         by one of the assembler files found in asm.  Look at des/asm/readme
  58         for more info.
  59
  60         /* Modification to fcrypt so it can be compiled to support
  61         HPUX 10.x's long password format, define -DLONGCRYPT to use this.
  62         Thanks to Jens Kupferschmidt <bt1cu@hpboot.rz.uni-leipzig.de>. */
  63
  64         SIGWINCH case put in des_read_passwd() so the function does not
  65         'exit' if this function is recieved.
  66
  67 Version 3.25 17/07/96
  68         Modified read_pwd.c so that stdin can be read if not a tty.
  69         Thanks to Jeff Barber <jeffb@issl.atl.hp.com> for the patches.
  70         des_init_random_number_generator() shortened due to VMS linker
  71         limits.
  72         Added RSA's DESX cbc mode.  It is a form of cbc encryption, with 2
  73         8 byte quantites xored before and after encryption.
  74         des_xcbc_encryption() - the name is funny to preserve the des_
  75         prefix on all functions.
  76
  77 Version 3.24 20/04/96
  78         The DES_PTR macro option checked and used by SSLeay configuration
  79
  80 Version 3.23 11/04/96
  81         Added DES_LONG.  If defined to 'unsigned int' on the DEC Alpha,
  82         it gives a %20 speedup :-)
  83         Fixed the problem with des.pl under perl5.  The patches were
  84         sent by Ed Kubaitis (ejk@uiuc.edu).
  85         if fcrypt.c, changed values to handle illegal salt values the way
  86         normal crypt() implementations do.  Some programs apparently use
  87         them :-(. The patch was sent by Bjorn Gronvall <bg@sics.se>
  88
  89 Version 3.22 29/11/95
  90         Bug in des(1), an error with the uuencoding stuff when the
  91         'data' is small, thanks to Geoff Keating <keagchon@mehta.anu.edu.au>
  92         for the patch.
  93
  94 Version 3.21 22/11/95
  95         After some emailing back and forth with
  96         Colin Plumb <colin@nyx10.cs.du.edu>, I've tweaked a few things
  97         and in a future version I will probably put in some of the
  98         optimisation he suggested for use with the DES_USE_PTR option.
  99         Extra routines from Mark Murray <mark@grondar.za> for use in
 100         freeBSD.  They mostly involve random number generation for use
 101         with kerberos.  They involve evil machine specific system calls
 102         etc so I would normally suggest pushing this stuff into the
 103         application and/or using RAND_seed()/RAND_bytes() if you are
 104         using this DES library as part of SSLeay.
 105         Redone the read_pw() function so that it is cleaner and
 106         supports termios, thanks to Sameer Parekh <sameer@c2.org>
 107         for the initial patches for this.
 108         Renamed 3ecb_encrypt() to ecb3_encrypt().  This has been
 109          done just to make things more consistent.
 110         I have also now added triple DES versions of cfb and ofb.
 111
 112 Version 3.20
 113         Damn, Damn, Damn, as pointed out by Mike_Spreitzer.PARC@xerox.com,
 114         my des_random_seed() function was only copying 4 bytes of the
 115         passed seed into the init structure.  It is now fixed to copy 8.
 116         My own suggestion is to used something like MD5 :-)
 117
 118 Version 3.19
 119         While looking at my code one day, I though, why do I keep on
 120         calling des_encrypt(in,out,ks,enc) when every function that
 121         calls it has in and out the same.  So I dropped the 'out'
 122         parameter, people should not be using this function.
 123
 124 Version 3.18 30/08/95
 125         Fixed a few bit with the distribution and the filenames.
 126         3.17 had been munged via a move to DOS and back again.
 127         NO CODE CHANGES
 128
 129 Version 3.17 14/07/95
 130         Fixed ede3 cbc which I had broken in 3.16.  I have also
 131         removed some unneeded variables in 7-8 of the routines.
 132
 133 Version 3.16 26/06/95
 134         Added des_encrypt2() which does not use IP/FP, used by triple
 135         des routines.  Tweaked things a bit elsewhere. %13 speedup on
 136         sparc and %6 on a R4400 for ede3 cbc mode.
 137
 138 Version 3.15 06/06/95
 139         Added des_ncbc_encrypt(), it is des_cbc mode except that it is
 140         'normal' and copies the new iv value back over the top of the
 141         passed parameter.
 142         CHANGED des_ede3_cbc_encrypt() so that it too now overwrites
 143         the iv.  THIS WILL BREAK EXISTING CODE, but since this function
 144         only new, I feel I can change it, not so with des_cbc_encrypt :-(.
 145         I need to update the documentation.
 146
 147 Version 3.14 31/05/95
 148         New release upon the world, as part of my SSL implementation.
 149         New copyright and usage stuff.  Basically free for all to use
 150         as long as you say it came from me :-)
 151
 152 Version 3.13 31/05/95
 153         A fix in speed.c, if HZ is not defined, I set it to 100.0
 154         which is reasonable for most unixes except SunOS 4.x.
 155         I now have a #ifdef sun but timing for SunOS 4.x looked very
 156         good :-(.  At my last job where I used SunOS 4.x, it was
 157         defined to be 60.0 (look at the old INSTALL documentation), at
 158         the last release had it changed to 100.0 since I now work with
 159         Solaris2 and SVR4 boxes.
 160         Thanks to  Rory Chisholm <rchishol@math.ethz.ch> for pointing this
 161         one out.
 162
 163 Version 3.12 08/05/95
 164         As pointed out by The Crypt Keeper <tck@bend.UCSD.EDU>,
 165         my D_ENCRYPT macro in crypt() had an un-necessary variable.
 166         It has been removed.
 167
 168 Version 3.11 03/05/95
 169         Added des_ede3_cbc_encrypt() which is cbc mode des with 3 keys
 170         and one iv.  It is a standard and I needed it for my SSL code.
 171         It makes more sense to use this for triple DES than
 172         3cbc_encrypt().  I have also added (or should I say tested :-)
 173         cfb64_encrypt() which is cfb64 but it will encrypt a partial
 174         number of bytes - 3 bytes in 3 bytes out.  Again this is for
 175         my SSL library, as a form of encryption to use with SSL
 176         telnet.
 177
 178 Version 3.10 22/03/95
 179         Fixed a bug in 3cbc_encrypt() :-(.  When making repeated calls
 180         to cbc3_encrypt, the 2 iv values that were being returned to
 181         be used in the next call were reversed :-(.
 182         Many thanks to Bill Wade <wade@Stoner.COM> for pointing out
 183         this error.
 184
 185 Version 3.09 01/02/95
 186         Fixed des_random_key to far more random, it was rather feeble
 187         with regards to picking the initial seed.  The problem was
 188         pointed out by Olaf Kirch <okir@monad.swb.de>.
 189
 190 Version 3.08 14/12/94
 191         Added Makefile.PL so libdes can be built into perl5.
 192         Changed des_locl.h so RAND is always defined.
 193
 194 Version 3.07 05/12/94
 195         Added GNUmake and stuff so the library can be build with
 196         glibc.
 197
 198 Version 3.06 30/08/94
 199         Added rpc_enc.c which contains _des_crypt.  This is for use in
 200         secure_rpc v 4.0
 201         Finally fixed the cfb_enc problems.
 202         Fixed a few parameter parsing bugs in des (-3 and -b), thanks
 203         to Rob McMillan <R.McMillan@its.gu.edu.au>
 204
 205 Version 3.05 21/04/94
 206         for unsigned long l; gcc does not produce ((l>>34) == 0)
 207         This causes bugs in cfb_enc.
 208         Thanks to Hadmut Danisch <danisch@ira.uka.de>
 209
 210 Version 3.04 20/04/94
 211         Added a version number to des.c and libdes.a
 212
 213 Version 3.03 12/01/94
 214         Fixed a bug in non zero iv in 3cbc_enc.
 215
 216 Version 3.02 29/10/93
 217         I now work in a place where there are 6+ architectures and 14+
 218         OS versions :-).
 219         Fixed TERMIO definition so the most sys V boxes will work :-)
 220
 221 Release upon comp.sources.misc
 222 Version 3.01 08/10/93
 223         Added des_3cbc_encrypt()
 224
 225 Version 3.00 07/10/93
 226         Fixed up documentation.
 227         quad_cksum definitely compatible with MIT's now.
 228
 229 Version 2.30 24/08/93
 230         Triple DES now defaults to triple cbc but can do triple ecb
 231          with the -b flag.
 232         Fixed some MSDOS uuen/uudecoding problems, thanks to
 233         Added prototypes.
 234
 235 Version 2.22 29/06/93
 236         Fixed a bug in des_is_weak_key() which stopped it working :-(
 237         thanks to engineering@MorningStar.Com.
 238
 239 Version 2.21 03/06/93
 240         des(1) with no arguments gives quite a bit of help.
 241         Added -c (generate ckecksum) flag to des(1).
 242         Added -3 (triple DES) flag to des(1).
 243         Added cfb and ofb routines to the library.
 244
 245 Version 2.20 11/03/93
 246         Added -u (uuencode) flag to des(1).
 247         I have been playing with byte order in quad_cksum to make it
 248          compatible with MIT's version.  All I can say is avid this
 249          function if possible since MIT's output is endian dependent.
 250
 251 Version 2.12 14/10/92
 252         Added MSDOS specific macro in ecb_encrypt which gives a %70
 253          speed up when the code is compiled with turbo C.
 254
 255 Version 2.11 12/10/92
 256         Speedup in set_key (recoding of PC-1)
 257          I now do it in 47 simple operations, down from 60.
 258          Thanks to John Fletcher (john_fletcher@lccmail.ocf.llnl.gov)
 259          for motivating me to look for a faster system :-)
 260          The speedup is probably less that 1% but it is still 13
 261          instructions less :-).
 262
 263 Version 2.10 06/10/92
 264         The code now works on the 64bit ETA10 and CRAY without modifications or
 265          #defines.  I believe the code should work on any machine that
 266          defines long, int or short to be 8 bytes long.
 267         Thanks to Shabbir J. Safdar (shabby@mentor.cc.purdue.edu)
 268          for helping me fix the code to run on 64bit machines (he had
 269          access to an ETA10).
 270         Thanks also to John Fletcher <john_fletcher@lccmail.ocf.llnl.gov>
 271          for testing the routines on a CRAY.
 272         read_password.c has been renamed to read_passwd.c
 273         string_to_key.c has been renamed to string2key.c
 274
 275 Version 2.00 14/09/92
 276         Made mods so that the library should work on 64bit CPU's.
 277         Removed all my uchar and ulong defs.  To many different
 278          versions of unix define them in their header files in too many
 279          different combinations :-)
 280         IRIX - Sillicon Graphics mods (mostly in read_password.c).
 281          Thanks to Andrew Daviel (advax@erich.triumf.ca)
 282
 283 Version 1.99 26/08/92
 284         Fixed a bug or 2 in enc_read.c
 285         Fixed a bug in enc_write.c
 286         Fixed a pseudo bug in fcrypt.c (very obscure).
 287
 288 Version 1.98 31/07/92
 289         Support for the ETA10.  This is a strange machine that defines
 290         longs and ints as 8 bytes and shorts as 4 bytes.
 291         Since I do evil things with long * that assume that they are 4
 292         bytes.  Look in the Makefile for the option to compile for
 293         this machine.  quad_cksum appears to have problems but I
 294         will don't have the time to fix it right now, and this is not
 295         a function that uses DES and so will not effect the main uses
 296         of the library.
 297
 298 Version 1.97 20/05/92 eay
 299         Fixed the Imakefile and made some changes to des.h to fix some
 300         problems when building this package with Kerberos v 4.
 301
 302 Version 1.96 18/05/92 eay
 303         Fixed a small bug in string_to_key() where problems could
 304         occur if des_check_key was set to true and the string
 305         generated a weak key.
 306
 307 Patch2 posted to comp.sources.misc
 308 Version 1.95 13/05/92 eay
 309         Added an alternative version of the D_ENCRYPT macro in
 310         ecb_encrypt and fcrypt.  Depending on the compiler, one version or the
 311         other will be faster.  This was inspired by
 312         Dana How <how@isl.stanford.edu>, and her pointers about doing the
 313         *(ulong *)((uchar *)ptr+(value&0xfc))
 314         vs
 315         ptr[value&0x3f]
 316         to stop the C compiler doing a <<2 to convert the long array index.
 317
 318 Version 1.94 05/05/92 eay
 319         Fixed an incompatibility between my string_to_key and the MIT
 320          version.  When the key is longer than 8 chars, I was wrapping
 321          with a different method.  To use the old version, define
 322          OLD_STR_TO_KEY in the makefile.  Thanks to
 323          viktor@newsu.shearson.com (Viktor Dukhovni).
 324
 325 Version 1.93 28/04/92 eay
 326         Fixed the VMS mods so that echo is now turned off in
 327          read_password.  Thanks again to brennan@coco.cchs.su.oz.AU.
 328         MSDOS support added.  The routines can be compiled with
 329          Turbo C (v2.0) and MSC (v5.1).  Make sure MSDOS is defined.
 330
 331 Patch1 posted to comp.sources.misc
 332 Version 1.92 13/04/92 eay
 333         Changed D_ENCRYPT so that the rotation of R occurs outside of
 334          the loop.  This required rotating all the longs in sp.h (now
 335          called spr.h). Thanks to Richard Outerbridge <71755.204@CompuServe.COM>
 336         speed.c has been changed so it will work without SIGALRM.  If
 337          times(3) is not present it will try to use ftime() instead.
 338
 339 Version 1.91 08/04/92 eay
 340         Added -E/-D options to des(1) so it can use string_to_key.
 341         Added SVR4 mods suggested by witr@rwwa.COM
 342         Added VMS mods suggested by brennan@coco.cchs.su.oz.AU.  If
 343         anyone knows how to turn of tty echo in VMS please tell me or
 344         implement it yourself :-).
 345         Changed FILE *IN/*OUT to *DES_IN/*DES_OUT since it appears VMS
 346         does not like IN/OUT being used.
 347
 348 Libdes posted to comp.sources.misc
 349 Version 1.9 24/03/92 eay
 350         Now contains a fast small crypt replacement.
 351         Added des(1) command.
 352         Added des_rw_mode so people can use cbc encryption with
 353         enc_read and enc_write.
 354
 355 Version 1.8 15/10/91 eay
 356         Bug in cbc_cksum.
 357         Many thanks to Keith Reynolds (keithr@sco.COM) for pointing this
 358         one out.
 359
 360 Version 1.7 24/09/91 eay
 361         Fixed set_key :-)
 362         set_key is 4 times faster and takes less space.
 363         There are a few minor changes that could be made.
 364
 365 Version 1.6 19/09/1991 eay
 366         Finally go IP and FP finished.
 367         Now I need to fix set_key.
 368         This version is quite a bit faster that 1.51
 369
 370 Version 1.52 15/06/1991 eay
 371         20% speedup in ecb_encrypt by changing the E bit selection
 372         to use 2 32bit words.  This also required modification of the
 373         sp table.  There is still a way to speedup the IP and IP-1
 374         (hints from outer@sq.com) still working on this one :-(.
 375
 376 Version 1.51 07/06/1991 eay
 377         Faster des_encrypt by loop unrolling
 378         Fixed bug in quad_cksum.c (thanks to hughes@logos.ucs.indiana.edu)
 379
 380 Version 1.50 28/05/1991 eay
 381         Optimised the code a bit more for the sparc.  I have improved the
 382         speed of the inner des_encrypt by speeding up the initial and
 383         final permutations.
 384
 385 Version 1.40 23/10/1990 eay
 386         Fixed des_random_key, it did not produce a random key :-(
 387
 388 Version 1.30  2/10/1990 eay
 389         Have made des_quad_cksum the same as MIT's, the full package
 390         should be compatible with MIT's
 391         Have tested on a DECstation 3100
 392         Still need to fix des_set_key (make it faster).
 393         Does des_cbc_encrypts at 70.5k/sec on a 3100.
 394
 395 Version 1.20 18/09/1990 eay
 396         Fixed byte order dependencies.
 397         Fixed (I hope) all the word alignment problems.
 398         Speedup in des_ecb_encrypt.
 399
 400 Version 1.10 11/09/1990 eay
 401         Added des_enc_read and des_enc_write.
 402         Still need to fix des_quad_cksum.
 403         Still need to document des_enc_read and des_enc_write.
 404
 405 Version 1.00 27/08/1990 eay
 406